Pith. sign in

REVIEW 4 major objections 5 minor 55 references

Spectral Analysis of Node- and Cell-Centered Higher-Order Compact Schemes for Fully Discrete One and Two-Dimensional Convection-Dispersion Equation

T0 review · 4 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

Pith's one-line read Global spectral analysis fixes the critical dispersion numbers of compact convection-dispersion schemes at $D_{\alpha,\mathrm{cr}}=0.12$ (node-centered) and $0.012$ (cell-centered), beyond which spurious waves and reversed energy…

desk verdict Useful GSA maps for three compact schemes on the linear KdV model, but the headline stability thresholds are not actually pinned down by the figures as printed. read the letter →

arxiv 2507.14692 v1 pith:XO4JRC6I submitted 2025-07-19 math.NA cs.NA

classification math.NAcs.NA MSC 35Q5365M0665M12
keywords globalspectralanalysisconvection-dispersionequationKorteweg-deVriescompactfinitedifferenceschemesSSPRK3amplificationfactorq-wavesdispersionnumber
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper establishes sharp stability limits for three fully discrete compact schemes on the convection-dispersion equation, the linear prototype behind the KdV and mKdV equations. Coupled with SSPRK3 time stepping, the sixth- and eighth-order node-centered schemes CNCS6 and CNCS8 remain stable up to a dispersion number $D_\alpha\approx0.11$ over the full Nyquist range, with $D_{\alpha,\mathrm{cr}}=0.12$ as the critical value; the eighth-order cell-centered CCS8 has $D_{\alpha,\mathrm{cr}}=0.012$, an order of magnitude smaller, in exchange for resolving wavenumbers up to $kh=2\pi$. Beyond these thresholds the numerical amplification factor $|G_{\rm num}|$ exceeds unity, and the normalized phase speed and group velocity develop negative regions that the paper identifies as spurious q-waves, phase reversal, and reversed energy transport. Numerical tests on linear 1D/2D problems, KdV single and double solitons, small-dispersion KdV, and mKdV interactions confirm that instability appears at the predicted critical dispersion numbers and that the cell-centered scheme is markedly more accurate but computationally dearer.

What carries the argument

The load-bearing object is the numerical amplification factor of the fully discrete scheme, $G_{\rm num}=1-A+\tfrac{A^2}{2}-\tfrac{A^3}{6}$, with $A=N_c\, i k_{\rm eq}^{[1]}h + D_\alpha\big(-i (k_{\rm eq}^{[2]})^3 h^3\big)$ in 1D and the obvious four-term sum in 2D. The factors $k_{\rm eq}^{[1]}$ and $k_{\rm eq}^{[2]}$ are the equivalent wavenumbers of the first- and third-derivative compact stencils, rational trigonometric functions of $kh$ listed in Appendix B; they convert the physical dispersion relation $\omega=ck-\alpha k^3$ into the numerical relation used to plot $c_{\rm num}/c_{\rm ph}$ and $v_{g,\rm num}/v_g$. Stability is decided by whether $|G_{\rm num}|$ exceeds unity anywhere in the $(N_c,kh)$ plane, and nonphysical behavior is read from sign changes in the two velocity ratios, with the spectral domain extended to $kh\in[0,2\pi]$ for CCS8 to exploit its cell-centered degrees of freedom. The 2D analysis folds in the mesh aspect ratio $AR$ and wave angle $\theta$ through direction-wise Courant and dispersion numbers, fixing $AR=1$ and $\theta=45^\circ$ for the reported contours.

What would settle it

Evaluate $|G_{\rm num}|$ from Eq. (27) using the equivalent-wavenumber formulas of Appendix B at $N_c=1.4$ and $D_\alpha=0.12$ over $kh\in[0,\pi]$: the claimed threshold stands only if some wavenumber gives $|G_{\rm num}|>1$; if none does, the critical dispersion number for CNCS6/CNCS8 is misidentified. Alternatively, run Example 4.2 at $D_\alpha=0.115$ and $D_\alpha=0.125$ and compare the evolution of the maximum amplitude, since the paper states the nonlinear blow-up results themselves are not presented.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that global spectral analysis of the fully discrete linear equation $u_t+c u_x+\alpha u_{xxx}=0$ (with the 2D extension $u_t+c_xu_x+c_yu_y+\alpha(u_{xxx}+u_{yyy})=0$) identifies dispersion numbers $D_\alpha=\alpha\Delta t/h^3$ at which the schemes change character: $D_{\alpha,\mathrm{cr}}=0.12$ for both CNCS6 and CNCS8, and $D_{\alpha,\mathrm{cr}}=0.012$ for CCS8. Below these values, with Courant numbers $N_c\lesssim1.3$ for the node-centered schemes and $N_c\lesssim0.65$ for CCS8, $|G_{\rm num}|\le1$ over the full resolved spectrum, and the numerical phase speed and group velocity ratios stay close to unity over a wide wavenumber range. At the critical values, stability is lost at high wavenumbers ($kh\approx2.5$ for CNCS, $kh\approx5.35$ for CCS8), and the contour plots of $c_{\rm num}/c_{\rm ph}$ and $v_{g,\rm num}/v_g$ show negative regions where waves travel backward and energy is transported opposite to the physical direction, which the paper calls q-waves. The experiments in Section 4 are reported as validating these thresholds: linear and nonlinear KdV-type simulations blow up at $D_{\alpha,\mathrm{cr}}$, and CCS8 systematically produces errors about an order of magnitude smaller than CNCS8 at at least double the computational cost.

Load-bearing premise

The load-bearing premise is that the linear, constant-coefficient, periodic-grid thresholds computed from Eq. (13) also govern the nonlinear KdV and mKdV simulations, so that choosing CFL=0.11 for CNCS and 0.011 for CCS and expecting blow-up at $D_{\alpha,\mathrm{cr}}=0.12$ and $0.012$ follows from the linear analysis; the paper explicitly notes that the nonlinear blow-up results are not presented.

Editorial extensions

If this is right

  • For CNCS6 and CNCS8 with SSPRK3, keeping $D_\alpha\le0.11$ and $N_c\lesssim1.3$ avoids $|G_{\rm num}|>1$ over the complete $kh\in[0,\pi]$ range; at $D_{\alpha,\mathrm{cr}}=0.12$ the usable wavenumber range drops to about $kh\le2.5$.
  • For CCS8, $D_\alpha$ must be kept near $0.011$–$0.012$ and $N_c$ below about $0.65$, but the scheme then resolves the doubled range $kh\in[0,2\pi]$, making it the choice when high-wavenumber fidelity matters more than large time steps.
  • Spurious q-waves, phase reversal, and reversed energy transport can occur even where $|G_{\rm num}|\le1$, so parameter selection should be guided by the normalized phase speed and group velocity contours, not by stability alone.
  • The same critical dispersion numbers carry over to the 2D linear tests at $AR=1$, $\theta=45^\circ$; CNCS8 shrinks the negative-group-velocity area relative to CNCS6 but does not eliminate it.
  • CCS8's accuracy advantage is about an order of magnitude in $L_\infty$ error over CNCS8 at similar resolution, with convergence orders near 8 for CNCS schemes and near 7 for CCS8, at a runtime cost at least double that of CNCS.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The transfer of linear periodic-grid thresholds to the nonlinear KdV/mKdV runs is asserted, not demonstrated: the paper says the nonlinear blow-up results are not presented. A direct check is to run the single-soliton example at $D_\alpha=0.115$ and $0.125$ and record whether the solution remains bounded near the predicted edge.
  • The same $G_{\rm num}$ machinery can be applied at other aspect ratios and incidence angles; the reported $AR=1$, $\theta=45^\circ$ slice leaves open the question of how directional anisotropy shifts $D_{\alpha,\mathrm{cr}}$ in the 2D case.
  • Because the q-wave and phase-reversal regions appear inside the stable zone, the contours of $c_{\rm num}/c_{\rm ph}$ and $v_{g,\rm num}/v_g$ could be used as local error estimators for adaptive stepping or grid refinement in KdV-type solvers, an extension the paper does not pursue.
  • The claimed doubled spectral resolution of CCS8 can be tested directly with a single-mode linear problem at $kh\in(\pi,2\pi)$: the observed phase error should match the paper's equivalent-wavenumber curves and stay small where those curves stay near unity.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper presents a global spectral analysis (GSA) of the fully discrete one- and two-dimensional linear convection–dispersion equation, using three high-order compact spatial discretizations (CNCS6, CNCS8, CCS8) combined with SSPRK3 time integration. The authors derive the numerical amplification factor Gnum, normalized phase speed, and normalized group velocity as functions of wavenumber, Courant number Nc, and dispersion number Dα, and use these to identify critical dispersion numbers Dα,cr and Courant limits beyond which |Gnum| exceeds unity, with associated nonphysical features such as q-waves, phase reversal, and reversed energy transport. The theoretical predictions are compared with numerical experiments for linear problems, KdV and mKdV soliton problems, and small-dispersion KdV shock problems.

Significance. If the claimed thresholds are correct, the paper offers practically useful guidance for selecting discretization parameters for convection–dispersion simulations. The derivation of Gnum in Eq. (27) from the SSPRK3 polynomial is standard, the equivalent wavenumber expressions in Appendix B are explicitly given, and the convergence tables in Section 4 are consistent with the nominal orders of the schemes. The paper also makes a useful comparison between node-centered and cell-centered compact schemes. However, the central stability-threshold claims are not actually pinned down by the presented contour plots: the figure captions report maximum |Gnum| values far above unity for parameter values that the text describes as stable. In addition, the normalized phase and group velocity diagnostics divide by quantities that vanish on curves, so contour extrema of order 10^4–10^6 include singular features rather than physically meaningful growth. These issues affect the headline conclusions and require careful revision.

major comments (4)
  1. [§3.1.1–3.1.3 and Figs. 2–4] The central stability claim that CNCS6 is 'stable throughout the Nyquist range' at Dα=0.11 is contradicted by the caption of Fig. 2(i), which reports max |Gnum|=21.09 at that parameter value; similarly, Fig. 3(i) reports max |Gnum|=25.16 at Dα=0.11 for CNCS8, and Fig. 4(i) reports max |Gnum|=2.60e+02 at Dα=0.011 for CCS8. Since |Gnum|>1 is precisely the instability criterion used to define Dα,cr, the critical values Dα,cr=0.12 and 0.012 are not established unless the relevant (Nc,kh) window over which the maximum is taken is specified and the text is reconciled with the displayed maxima.
  2. [Eqs. (31)–(32) and (43)–(45)] The normalized phase speed cnum/cph and group velocity vg,num/vg divide by cph=c−αk² and vg=c−3αk², respectively, which vanish along curves in the (Nc,kh) plane; in nondimensional form the denominators are Nc−Dα(kh)² and Nc−3Dα(kh)². The contour extrema reported in Figs. 2–7, ranging up to 10^6, therefore include singular values along these curves, and the associated statements about expansion of 'negative q-wave' and 'reversed energy transport' regions are not meaningful until the analysis is restricted to the nonsingular set or a different normalization is used.
  3. [§4.1, Eq. (46), and Fig. 8] The numerical experiment in Example 4.1 uses N=100, h=2π/100, and the stated rule CFL=0.11 for CNCS; substituting into Eq. (46) with g'(u)=2 and f'(u)=1 gives Dα = CFL/(1+2h²) ≈ 0.109, so the claimed destabilizing run at Dα=0.12 cannot be produced with the stated CFL value. The paper needs to specify either how Dα=0.12 is obtained in practice or correct the time-step prescription, because the validation of Dα,cr depends on actually performing the run at the claimed critical value.
  4. [§4.2, Example 4.2] The statement that nonlinear blow-up occurs at Dα,cr=0.12 for CNCS and 0.012 for CCS, with 'these results are not presented here,' is an unsupported assertion. Since the GSA applies to the linear constant-coefficient equation, the transfer of the threshold to the nonlinear KdV/mKdV computations is load-bearing and needs direct evidence; nonlinearity and the filtering procedure used in Example 4.3 could plausibly shift the thresholds.
minor comments (5)
  1. [Abstract] The phrase 'Courant Friedrichs Lewy number' should be hyphenated as 'Courant–Friedrichs–Lewy number'.
  2. [§3.1.1] The text says that for Dα=0.11 the scheme 'remains stable throughout the Nyquist range (kh=π)', but the caption of Fig. 2(i) reports max |Gnum|=21.09; please clarify whether the contour plots are clipped or the stability statement refers to a restricted (Nc,kh) window.
  3. [§3.2.1] The critical value Dα,cr=0.12 in the 2D analysis is stated for a fixed Courant number (Nc=0.9 for CNCS6, Nc=0.7 for CNCS8, Nc=0.45 for CCS8), but the 1D analysis quotes thresholds over a range of Nc; the dependence of Dα,cr on Nc should be stated explicitly to avoid ambiguity.
  4. [Fig. 9] The caption of Fig. 9 does not identify the line styles or colors used for CNCS6, CNCS8, and CCS8; please add a legend or specify the correspondence in the caption.
  5. [§1, Introduction] The sentence 'To validate the theoretical findings, we validate our spectral predictions through numerical experiments' is repetitive; consider rewording to avoid the doubled 'validate'.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the GSA thresholds are direct evaluations of Gnum, not fitted constants or self-citation imports.

full rationale

The central derivation is self-contained. Equations (13)-(27) derive Gnum from the PDE, the discrete Fourier representation, and the SSPRK3 stability polynomial, with the equivalent wavenumbers in Appendix B obtained from the stencils stated in Sections 2.1.1-2.1.2. The critical dispersion numbers (Dα ≈ 0.12 for CNCS6/CNCS8 and ≈ 0.012 for CCS8) are read from the contour plots where |Gnum| exceeds unity; they are outputs of the paper's own spectral computation, not fitted parameters, and the numerical experiments do not feed back into the definition of Dα,cr. The use of the authors' CCS8 scheme and filter from reference [30] is not load-bearing circularity: the CCS8 coefficients and equivalent-wavenumber formulas are restated in the paper itself, so the analysis does not depend on unverified content of the self-citation. The numerical experiments are self-consistency tests of the same schemes rather than independent empirical confirmation, which is standard for spectral analysis and does not make the derivation circular. Concerns such as the sentence in Example 4.2 that blow-up occurs at the critical values 'although these results are not presented here', and the apparent inconsistency that Fig. 2 reports max |Gnum| = 21.09 at the claimed stable Dα = 0.11, are correctness and rigor issues about an under-specified Nc window and missing experiment data, not circularity. No step reduces a prediction to an input by construction.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The analysis relies on standard Fourier and Laplace representation, the use of the linear constant-coefficient convection-dispersion equation as a surrogate for nonlinear KdV and mKdV stability, and the claim that CCS8's combined node and cell-centered unknowns allow kh up to 2π. No parameter is fitted to output; the critical Dα values are computed, not calibrated. The CCS8 scheme and its filtering step come from the authors' prior arXiv preprint [30].

assumptions (4)
  • standard math The solution can be represented by Fourier-Laplace integrals with periodic boundary conditions on a uniform grid, and linear superposition holds for the model equation (Eqs. 14 and 17).
    This is the standard basis of GSA as used in [16,38,48]; it requires a linear constant-coefficient operator.
  • domain assumption The linear constant-coefficient equation Eq. (13) is the correct model for stability and dispersion of the nonlinear KdV and mKdV problems simulated in Section 4.
    Section 4 sets CFL based on linear GSA thresholds, and Example 4.2 asserts nonlinear blow-up at Dα,cr without showing results.
  • domain assumption CCS8's combination of node and cell-center unknowns extends the resolved wavenumber domain to kh = 2π.
    This is invoked in Sections 3.1.3 and 3.2.3 to justify evaluating CCS8 over [0,2π]; if only a single staggered field is considered, the Nyquist limit would be kh = π.
  • domain assumption The filtering procedure from [30] removes spurious oscillations without changing the essential wave dynamics in Example 4.3.
    Used to clean the discontinuous-initial-condition solutions; the filter's effect on the claimed validation is not analyzed.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Spectral Analysis of Node- and Cell-Centered Higher-Order Compact Schemes for Fully Discrete One and Two-Dimensional Convection-Dispersion Equation." pith.science (2026). https://pith.science/paper/XO4JRC6I

@misc{pith2026250714692,
  author       = {Pith},
  title        = {Pith review of: Spectral Analysis of Node- and Cell-Centered Higher-Order Compact Schemes for Fully Discrete One and Two-Dimensional Convection-Dispersion Equation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XO4JRC6I}},
  note         = {Machine review of arXiv:2507.14692}
}
abstract

In this study, we present a comprehensive global spectral analysis of the convection dispersion equation, which is also referred to in specific contexts as the Korteweg de Vries (KdV) equation, to investigate the behaviour of high order numerical schemes across a wide range of nondimensional parameters. The motivation for this analysis stems from the equation's importance in modeling wave propagation and transport phenomena, where accurate resolution of dispersive effects is critical, and traditional numerical schemes often suffer from spurious artifacts. We analyze one sixth order and two eighth order compact spatial discretization schemes, encompassing both node centered and cell centered formulations, combined with a third order strong stability preserving Runge Kutta (SSPRK3) time integrator. The analysis is performed in terms of key nondimensional parameters such as the wavenumber, Courant Friedrichs Lewy number $N_c$, and dispersion number $D_{\alpha}$ over the full spectral plane for both one and two dimensional cases. Key numerical indicators, including the amplification factor, normalized phase speed, and normalized group velocity, are evaluated to characterize stability, dispersion error, errors in energy transport, and directional anisotropy. Critical dispersion thresholds and Courant numbers are identified, beyond which numerical instability and nonphysical phenomena such as spurious q waves and reversed phase or energy transport arise. Theoretical predictions are validated through numerical experiments involving linear and nonlinear one and two dimensional test problems, including cases with exact solutions and established benchmark results. This comprehensive analysis uncovers subtle numerical errors and offers practical guidance for selecting reliable discretization parameters, ensuring accurate and stable simulations of convection dispersion systems.

Figures

Figures reproduced from arXiv: 2507.14692 by the authors.

Figure 1
Figure 1. Spectral resolution characteristics of indicated first- and [PITH_FULL_IMAGE:figures/full_fig_p006_1.png] view at source ↗
Figure 2
Figure 2. Contour plots of |Gnum|, cnum cph , and vg,num vg in the (Nc, kh) plane for Eq. (13), computed using the SSPRK3-CNCS6 scheme with the indicated values of Dα. 12 [PITH_FULL_IMAGE:figures/full_fig_p012_2.png] view at source ↗
Figure 3
Figure 3. Contour plots of |Gnum|, cnum cph , and vg,num vg in the (Nc, kh) plane for Eq. (13), computed using the SSPRK3-CNCS8 scheme with the indicated values of Dα. 14 [PITH_FULL_IMAGE:figures/full_fig_p014_3.png] view at source ↗
Figures from the paper (15 more)
Figure 4
Figure 4. Figure 4: Contour plots of |Gnum|, cnum cph , and vg,num vg in the (Nc, kh) plane for Eq. (13), computed using the SSPRK3-CCS8 scheme with the indicated values of Dα. 15 [PITH_FULL_IMAGE:figures/full_fig_p015_4.png]
Figure 5
Figure 5. Figure 5: Contour plots of |Gnum|, cnum cph , (vg,x)num vg,x in the (kxhx, kyhy)− plane for Eq. (33) with Nc = 0.9, computed using the SSPRK3-CNCS6 scheme with the indicated values of Dα. 19 [PITH_FULL_IMAGE:figures/full_fig_p019_5.png]
Figure 6
Figure 6. Figure 6: Contour plots of |Gnum|, cnum cph , (vg,x)num vg,x in the (kxhx, kyhy)− plane for Eq. (33) with Nc = 0.7, computed using the SSPRK3-CNCS8 scheme with the indicated values of Dα. 20 [PITH_FULL_IMAGE:figures/full_fig_p020_6.png]
Figure 7
Figure 7. Figure 7: Contour plots of |Gnum|, cnum cph , (vg,x)num vg,x in the (kxhx, kyhy)− plane for Eq. (33) with Nc = 0.45, computed using the SSPRK3-CCS8 scheme with the indicated values of Dα. 22 [PITH_FULL_IMAGE:figures/full_fig_p022_7.png]
Figure 8
Figure 8. Figure 8: Variation of the maximum absolute difference between the exa [PITH_FULL_IMAGE:figures/full_fig_p024_8.png]
Figure 9
Figure 9. Figure 9: Evolution of the maximum error over time for example 4.1 with th [PITH_FULL_IMAGE:figures/full_fig_p024_9.png]
Figure 10
Figure 10. Figure 10: Solutions and errors for the initial condition (50) of example [PITH_FULL_IMAGE:figures/full_fig_p026_10.png]
Figure 11
Figure 11. Figure 11: Solutions and errors for the initial condition (51) of example [PITH_FULL_IMAGE:figures/full_fig_p027_11.png]
Figure 12
Figure 12. Figure 12: Numerical solutions for initial condition (53) of example 4.3. [PITH_FULL_IMAGE:figures/full_fig_p028_12.png]
Figure 13
Figure 13. Figure 13: Numerical solutions for initial condition (54) of example 4.3. [PITH_FULL_IMAGE:figures/full_fig_p028_13.png]
Figure 14
Figure 14. Figure 14: Solutions and errors for the initial condition (57) of example [PITH_FULL_IMAGE:figures/full_fig_p029_14.png]
Figure 15
Figure 15. Figure 15: Numerical solutions for initial condition (58) of example 4.4. [PITH_FULL_IMAGE:figures/full_fig_p030_15.png]
Figure 17
Figure 17. Figure 17: 0 0.1 0.2 0.3 0.4 0.5 Time 1084 10184 10284 (a) SSPRK3-CNCS6 0 0.1 0.2 0.3 0.4 0.5 Time 10-16 1084 10184 10284 Maximum Amplitude (b) SSPRK3-CNCS8 0 0.1 0.2 0.3 0.4 0.5 Time 1084 10184 10284 Maximum Amplitude (c) SSPRK3-CCS8 [PITH_FULL_IMAGE:figures/full_fig_p031_17.png]
Figure 16
Figure 16. Figure 16: Variation of the maximum amplitude of u over time for example 4.5. 5. Conclusions In this article, we have performed the GSA for the 1D and 2D convection-dispersion equations. For spatial discretization, we have used three compact schemes, CNCS6, CNCS8 and CCS8 with S…
Figure 17
Figure 17. Figure 17: Solutions and errors for example 4.5. The first row shows ex [PITH_FULL_IMAGE:figures/full_fig_p032_17.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

55 extracted references · 55 canonical work pages

  1. [27]

    Ashwin, K

    V. Ashwin, K. Saurabh, M. Sriramkrishnan, P. Bagade, M. Parv athi, T. K. Sengupta, Kdv equation and computations of solitons: Nonlinear error dynamic s, Journal of Sci- entific Computing 62 (3) (2015) 693–717

  2. [1]

    M. Guo, C. Fu, Y. Zhang, J. Liu, H. Yang, Study of ion-acoustic s olitary waves in a magnetized plasma using the three-dimensional time-space fract ional schamel-kdv equation, Complexity 2018 (1) (2018) 6852548

  3. [2]

    C. K. Tam, J. C. Webb, Dispersion-relation-preserving finite diffe rence schemes for computational acoustics, Journal of computational physics 107 (2) (1993) 262–281

  4. [3]

    M. R. Visbal, D. V. Gaitonde, Very high-order spatially implicit schem es for compu- tational acoustics on curvilinear meshes, Journal of Computation al Acoustics 9 (04) (2001) 1259–1286

  5. [4]

    M. H. Rahman, N. A. Chowdhury, A. Mannan, A. Mamun, Dust-ac oustic rogue waves in an electron-positron-ion-dust plasma medium, Galaxies 9 (2) (202 1) 31

  6. [5]

    D. A. Iordache, A. Petrescu, V. Iordache, Study of the solito ns propagation through optical fibers, UPB Scientific Bulletin A 72 (1) (2010) 53–58

  7. [6]

    Hasegawa, F

    A. Hasegawa, F. Tappert, Transmission of stationary nonlinear optical pulses in dis- persive dielectric fibers. ii. normal dispersion, Applied Physics Lette rs 23 (4) (1973) 171–172

  8. [7]

    D. J. Korteweg, G. De Vries, Xli. on the change of form of long wav es advancing in a rectangular canal, and on a new type of long stationary waves, The London, Edinburgh, and Dublin Philosophical Magazine and Journal of Science 39 (240) (1 895) 422–443

Show all 55 references
  1. [8]

    P. G. Drazin, R. S. Johnson, Solitons: an introduction, Vol. 2, Ca mbridge university press, 1989

  2. [9]

    Wazwaz, Partial differential equations and solitary waves t heory, Springer Science & Business Media, 2010

    A.-M. Wazwaz, Partial differential equations and solitary waves t heory, Springer Science & Business Media, 2010

  3. [10]

    Tanaka, Modified korteweg-de vries equation and scatterin g theory, Proceedings of the Japan Academy 48 (7) (1972) 466–469

    S. Tanaka, Modified korteweg-de vries equation and scatterin g theory, Proceedings of the Japan Academy 48 (7) (1972) 466–469

  4. [11]

    Khater, O

    A. Khater, O. El-Kalaawy, D. Callebaut, B¨ acklund transforma tions and exact solutions foralfv´ en solitons in a relativistic electron–positronplasma, Physica Scripta 58 (6) (1998) 545

  5. [12]

    T. S. Komatsu, S.-i. Sasa, Kink soliton characterizing traffic con gestion, Physical Review E 52 (5) (1995) 5574

  6. [13]

    H. Song, H. Ge, F. Chen, R. Cheng, Tdgl and mkdv equations fo r car-following model considering traffic jerk and velocity difference, Nonlinear Dynamics 8 7 (2017) 1809–1817

  7. [14]

    Helal, Soliton solution of some nonlinear partial differential equ ations and its appli- cations in fluid mechanics, Chaos, Solitons & Fractals 13 (9) (2002) 1 917–1929

    M. Helal, Soliton solution of some nonlinear partial differential equ ations and its appli- cations in fluid mechanics, Chaos, Solitons & Fractals 13 (9) (2002) 1 917–1929. 35

  8. [15]

    Leblond, D

    H. Leblond, D. Mihalache, Few-optical-cycle solitons: Modified ko rteweg–de vries sine- gordon equation versus other non–slowly-varying-envelope-approximation models, Phys- ical Review A—Atomic, Molecular, and Optical Physics 79 (6) (2009) 0 63835

  9. [16]

    Sengupta, High accuracy computing methods: fluid flows and wave phenomena, Cambridge University Press, 2013

    T. Sengupta, High accuracy computing methods: fluid flows and wave phenomena, Cambridge University Press, 2013

  10. [17]

    Hirsch, Numerical computation of internal and external flo ws, Computational meth- ods for inviscid and viscous flows 2

    C. Hirsch, Numerical computation of internal and external flo ws, Computational meth- ods for inviscid and viscous flows 2

  11. [18]

    Karunakar, S

    P. Karunakar, S. Chakraverty, Differential quadrature met hod for solving fifth-order kdv equations, in: Recent Trends in Wave Mechanics and Vibrations: Sele ct Proceedings of WMVC 2018, Springer, 2019, pp. 361–369

  12. [19]

    M. J. Ablowitz, P. A. Clarkson, Solitons, nonlinear evolution equa tions and inverse scattering, Vol. 149, Cambridge university press, 1991

  13. [20]

    A.-M. Wazwaz, The variational iteration method for rational so lutions for kdv, k (2, 2), burgers, and cubic boussinesq equations, Journal of Comput ational and Applied Mathematics 207 (1) (2007) 18–23

  14. [21]

    Z. Fu, S. Liu, S. Liu, New solutions to mkdv equation, Physics Let ters A 326 (5-6) (2004) 364–374

  15. [22]

    Levy, C.-W

    D. Levy, C.-W. Shu, J. Yan, Local discontinuous galerkin metho ds for nonlinear disper- sive equations, Journal of Computational Physics 196 (2) (2004) 751–772

  16. [23]

    Saucez, A

    P. Saucez, A. V. Wouwer, W. Schiesser, P. Zegeling, Method of lines study of nonlinear dispersive waves, Journal of Computational and Applied Mathemat ics 168 (1-2) (2004) 413–423

  17. [24]

    Ahmat, J

    M. Ahmat, J. Qiu, Direct weno scheme for dispersion-type equa tions, Mathematics and Computers in Simulation 204 (2023) 216–229

  18. [25]

    L. V. Salian, R. Samala, Exponential approximation space recon struction weighted es- sentially nonoscillatory scheme for dispersive partial differential eq uations, Mathemati- cal Methods in the Applied Sciences 47 (4) (2024) 1823–1851

  19. [26]

    J. Li, M. R. Visbal, High-order compact schemes for nonlinear dis persive waves, Journal of Scientific Computing 26 (2006) 1–23

  20. [28]

    S. K. Lele, Compact finite difference schemes with spectral-like r esolution, Journal of computational physics 103 (1) (1992) 16–42

  21. [29]

    L. V. Salian, S. Rathan, R. Kumar, Compact finite-difference sc heme for some sobolev type equations with dirichlet boundary conditions, arXiv preprint ar Xiv:2411.18445. 36

  22. [30]

    L. V. Salian, S. Rathan, D. Ghosh, A novel central compact fin ite-difference scheme for third derivatives with high spectral resolution, arXiv preprint arXiv :2405.00569

  23. [31]

    Suman, T

    V. Suman, T. K. Sengupta, C. J. D. Prasad, K. S. Mohan, D. Sa nwalia, Spectral analysis of finite difference schemes for convection diffusion equation, Comp uters & Fluids 150 (2017) 95–114

  24. [32]

    Sengupta, T

    S. Sengupta, T. K. Sengupta, J. K. Puttam, K. S. Vajjala, Glo bal spectral analysis for convection-diffusion-reaction equation in one and two-dimensions: Effects of numerical anti-diffusion and dispersion, Journal of Computational Physics 40 8 (2020) 109310

  25. [33]

    Sharma, A

    N. Sharma, A. Sengupta, M. Rajpoot, R. J. Samuel, T. K. Seng upta, Hybrid sixth order spatial discretization scheme for non-uniform cartesian grids, Co mputers & Fluids 157 (2017) 208–231

  26. [34]

    T. K. Sengupta, A. Sengupta, A new alternating bi-diagonal co mpact scheme for non- uniform grids, Journal of Computational Physics 310 (2016) 1–25

  27. [35]

    P. K. Maurya, M. K. Rajpoot, V. S. Yadav, A. Singh, New hybrid compact schemes for structured irregular meshes using birkhoff polynomial basis, Journ al of Computational Physics 423 (2020) 109808

  28. [36]

    L. N. Trefethen, Group velocity in finite difference schemes, SI AM review 24 (2) (1982) 113–136

  29. [37]

    Sagaut, V

    P. Sagaut, V. Suman, P. Sundaram, M. Rajpoot, Y. Bhumkar, S. Sengupta, A. Sen- gupta, T. Sengupta, Global spectral analysis: Review of numerica l methods, Computers & Fluids 261 (2023) 105915

  30. [38]

    T. K. Sengupta, A. Dipankar, P. Sagaut, Error dynamics: bey ond von neumann analysis, Journal of Computational Physics 226 (2) (2007) 1211–1218

  31. [39]

    P. K. Maurya, V. S. Yadav, B. Mahato, N. Ganta, M. K. Rajpoo t, Y. G. Bhumkar, New optimized implicit-explicit runge-kutta methods with applications to th e hyperbolic conservation laws, Journal of Computational Physics 446 (2021) 110650

  32. [40]

    T. K. Sengupta, A. Bhole, Error dynamics of diffusion equation: Effects of numerical diffusion and dispersive diffusion, Journal of Computational Physics 266 (2014) 240–251

  33. [41]

    Sengupta, N

    S. Sengupta, N. Sreejith, P. Mohanamuraly, G. Staffelbach, L . Gicquel, Global spectral analysis of the lax–wendroff-central difference scheme applied to c onvection–diffusion equation, Computers & Fluids 242 (2022) 105508

  34. [42]

    V. S. Yadav, N. Ganta, B. Mahato, M. K. Rajpoot, Y. G. Bhumk ar, New time-marching methods for compressible navier-stokes equations with application s to aeroacoustics problems, Applied Mathematics and Computation 419 (2022) 126863

  35. [43]

    V. S. Yadav, V. Maurya, M. K. Rajpoot, J. Jaglan, Spatiotemp oral pattern formations in stiff reaction-diffusion systems by new time marching methods, App lied Mathematics and Computation 431 (2022) 127299. 37

  36. [44]

    X. Liu, S. Zhang, H. Zhang, C.-W. Shu, A new class of central co mpact schemes with spectral-like resolution i: Linear schemes, Journal of Computation al Physics 248 (2013) 235–256

  37. [45]

    M. K. Rajpoot, S. Bhaumik, T. K. Sengupta, Solution of linearize d rotating shallow water equations by compact schemes with different grid-staggerin g strategies, Journal of Computational Physics 231 (5) (2012) 2300–2327

  38. [46]

    Gottlieb, C.-W

    S. Gottlieb, C.-W. Shu, Total variation diminishing runge-kutta s chemes, Mathematics of computation 67 (221) (1998) 73–85

  39. [47]

    Gottlieb, D

    S. Gottlieb, D. Ketcheson, C.-W. Shu, Strong stability preserv ing Runge-Kutta and multistep time discretizations, World Scientific, 2011

  40. [48]

    T. K. Sengupta, G. Ganeriwal, S. De, Analysis of central and up wind compact schemes, Journal of Computational Physics 192 (2) (2003) 677–694

  41. [49]

    Vichnevetsky, J

    R. Vichnevetsky, J. B. Bowles, Fourier analysis of numerical ap proximations of hyper- bolic equations, SIAM, 1982

  42. [50]

    T. K. Sengupta, Y. G. Bhumkar, M. K. Rajpoot, V. Suman, S. S aurabh, Spurious waves in discrete computation of wave phenomena and flow problems, Applie d Mathematics and Computation 218 (18) (2012) 9035–9065

  43. [51]

    T. K. Sengupta, M. K. Rajpoot, S. Saurabh, V. Vijay, Analysis of anisotropy of numeri- cal wave solutions by high accuracy finite difference methods, Jour nal of Computational Physics 230 (1) (2011) 27–60

  44. [52]

    Wazwaz, An analytic study on the third-order dispersive p artial differential equa- tions, Applied Mathematics and Computation 142 (2-3) (2003) 511– 520

    A.-M. Wazwaz, An analytic study on the third-order dispersive p artial differential equa- tions, Applied Mathematics and Computation 142 (2-3) (2003) 511– 520

  45. [53]

    Yoneyama, The korteweg-de vries two-soliton solution as int eracting two single soli- tons, Progress of theoretical physics 71 (4) (1984) 843–846

    T. Yoneyama, The korteweg-de vries two-soliton solution as int eracting two single soli- tons, Progress of theoretical physics 71 (4) (1984) 843–846

  46. [54]

    Ismail, F

    M. Ismail, F. Alotaibi, Numerical simulation of modified kortweg-de vries equation by linearized implicit schemes, Applied Mathematics 11 (11) (2020) 1139– 1161

  47. [55]

    Biswas, K

    A. Biswas, K. Raslan, Numerical simulation of the modified kortew eg-de vries equation, Physics of Wave Phenomena 19 (2) (2011) 142–147. 38

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.